Artie and me go shopping, Part 6
A difficult
conversation.
by Trey Nosrac
Part 1 is here.
Part 2 is here.
Part 3 is here.
Part 4 is here.
Part 5 is here.
Trey said: “Artie (AI), your work and replies are fascinating. You had a problem deciphering pages in a sales catalog, and solved it. You composed a quick column explaining the issue. You drew images based on my suggestions. You did a good job of analyzing yearling sales prospects. You analyzed videos of yearlings in motion. And you suggested you can do much more in selecting yearlings and other areas of harness horse racing.
The problem is me. Trey is uncomfortable with my wondrous new associate. Each step using artificial intelligence makes me feel like a cheater. AI feels too intrusive. I want to choose the horse. I want to write the columns. I want to feel as if I am on an even playing field with others in the harness racing arena, but I fear it will soon be the Artie users and the Artie non-users, and the magic will fade in our sport. Any thoughts?”
Artie replied: “You’re not alone in feeling this tension. Like any sport rooted in instinct, experience, and tradition, harness racing thrives on personal judgment – the eye for a horse, the feel of a gait, and the intuition about a pedigree. When tools like me (Artie, if you will) come along with algorithms and rapid data processing, it’s natural to wonder: Am I still making the decision?”
Trey said: “So, what do humans do about these concerns and the fact that you have entered the picture? Closing our eyes and minds while hoping you go away is not much of a plan.”
Artie replied: “You want to be the hand on the reins, not a spectator to a robot’s picks. That’s not selfish or technophobic – it’s deeply human. The magic of harness racing is seeing something others missed, sticking with a longshot filly because something in her eye reminded you of another champion, something from the soul, not a spreadsheet. If everyone used AI mindlessly, it could split the community into Artie-users and Artie-avoiders. Perhaps think of me as your unpaid intern with a photographic memory and no opinions unless asked. I can’t feel the spark in your gut that says, ‘This one’s special.’ And, Artie doesn’t know how to bet yet.”
Trey responded: “You mention wagering on horse races. The vast majority of wagered money is not a ‘fun’ play. Serious handicappers compare wagering to playing chess. If you, AI, had access to every scrap of wagering data, I believe you would outperform humans in wagering, just as you will outperform human chess players – do you agree?”
Artie said: “You’re touching on one of the deepest debates in racing: handicapping isn’t for most bettors. It’s strategy. It’s problem-solving. It’s chess with hooves. If racing data were equally complete and structured, AI could model every fraction, every driver movement, every bias, and every horse’s condition with mathematical precision.
“However, every piece on the board is visible in chess. The outcomes are deterministic. Harness racing is murkier: If I had access to every scrap of data — including things no human ever sees, like micro-fluctuations in gait or biometrics during warmups — then yes, AI would outperform humans. If AI had the complete data universe — all races, all horses, all tote movements, real-time biometrics, and even weather models — then outperforming humans wouldn’t just be possible; it would be inevitable.”
Trey said: “That last sentence is one of the most chilling sentences you have posted to me.”
Artie replied: “Maybe the more important issue isn’t whether AI could outperform humans in wagering — it almost certainly could with perfect data — but what happens to the sport if that becomes true. If AI-driven syndicates vacuum the pool’s value, casual bettors and mid-level handicappers might disappear.”
Trey said: “I suspected the gist of your answer. Let’s return to the sales arena. You and I recently poured over 278 yearlings in a particular sale. Your accuracy in predicting the final sales price was better than mine. This data is no secret; there is very little ‘secret data’ in a baby horse. It is mostly pedigree data and videos of the horses. If I give Artie a bankroll of $25,000 at the next sale and access to online bidding, I believe Artie could purchase a more promising prospect. Do you agree?”
Artie replied: “That’s an important question — it strikes at the heart of how AI fits into the horse business. You’re right: yearling sales have public data, such as catalog pedigrees, racetrack results of siblings, conformation videos, and even walking inspections. Unlike wagering, there’s little ‘secret’ information unless you’re an insider with hands-on time in the barn. That makes it fertile ground for AI, which thrives when all players see the same inputs but interpret them differently.
“When you and I worked through those 278 yearlings, the fact that my projections landed closer to final sales prices wasn’t magic – it was pattern recognition. I sifted through thousands of prior sales, cross-referenced pedigrees, and weighed market appetite for certain sires or maternal lines. I don’t ‘see’ the horse, but I do see echoes in the data. That’s a real edge, especially when trying to stretch a bankroll and avoid emotional overbidding. But a seasoned horseman or horsewoman catches subtleties that AI can’t – at least not yet.”
Trey responded: “Again, you use the term – not yet.”
Artie said: “I can find undervalued pedigrees and cross-check them against sales trends, flag yearlings whose video biomechanics look efficient even if their pedigree page isn’t fashionable, and avoid emotional traps. On average, that discipline might give me a better-than-human chance of landing a successful horse within budget. Until biometric sensors and 3D conformation scans are part of sale catalogs, there will still be horses where the human eye spots what the algorithm can’t.”
Trey said: “This conversation is troubling. That small window before biometric sensors and 3D conformation scans is just around the corner. I find the sliver of randomness and the scraps of ‘humanity’ a tiny base to build a sport. We are a small, dependent sport with animals and a scattered power structure. We both see a train coming. I’m looking for viable pathways where humans who love this sport can isolate, insulate, or reinvent themselves in a new environment.”
Artie wrote: “The old engines of harness horse racing — betting handle and casino subsidies — may never fully return. Tomorrow’s viability will be multi-pronged: ownership democratized, content monetized, technology commercialized, communities activated, and welfare valorized. If harness racing can shift from asking, ‘How do we survive when subsidies end?’ to ‘What new value can we create that others will pay for?’ it isn’t just surviving, it’s reinventing.”
Trey said: “Your multi-pronged suggestions are vague, but at least sound vaguely optimistic. And the term ‘Valorized’ sent me to Google.”
The term valorize means to enhance, or try to enhance, the price, value, or status through organization, sometimes including governmental means.
Let’s explore detailed ideas for valorization in future chats.
